H c: OAc

(9) Deulofeu, J . Chum. SOC., 2973 (1932). (10) Hockett, Dienes, Fletcher and Ramsden, TEIS JOUXNAL. 6S,. (1887). Wolfrom and Thompson, Tms JOURNAL,...
7 downloads 0 Views 449KB Size
957

STRUCTURE OF N-ACETYL-D-GLUCOFURANOSYLAMINE

June, 1‘34.1

[CONTRIBUTION FROM THE RESEARCH LABORATORY OF ORGANIC CHEMISTRY, MASSACHUSETTS INSTITUTE OF TECHNOLOGY, No. 2971

Lead Tetraacetate Oxidations in the Sugar Group. VIII.’ The Preparation and Proof of Structure of N-Acetyl-D-glucofuranosylamine2 BY ROBERTC. HOCKETT AND LEONARD B. CHANDLER* In 1893, Alfred W0h14 published his method of degrading an aldose to the next lower member of the series. The action of acetic anhydride and fused sodium acetate upon the D-glucose oxime which had been described by Jacobi6 produced an acetylated D-gluconic acid nitrile which in turn lost CN- when treated with an ammoniacal solution of silver oxide. This unusual hydrolysis of a cyanide group as a pseudohalogen instead of t o the carboxyl group, produced a compound with one less carbon atom. However, instead of obtaining D-arabinose directly in a free state, Wohl obtained the pentose in a combination with two molecules of acetamide. This product, usually called by the unsatisfactory name “D-arabinose diacetamide,” showed a composition and properties consistent with the formula HC(NHCOCH&

I I

HOCH HCOH HkOH

I

HzCOH

The open-chain structure was accepted as a matter of course in the era4 when both sugars and their oximes were regarded as compounds of this type. As the realization grew, however, that the majority of sugar derivatives have cyclic structures18 these acetamide condensation products eventually became recognized as representatives of the relatively rare class of acyclic sugar derivatives.? The absence of a ring has been further confirmed in several cases by observation of the number of acetyl groups introduced on complete a c e t y l a t i ~ nas ~ - well ~ as by study of the behavior of the substances toward 1ead.tetraacetate. lo (1) Number VI1 of this series, Taxa JOUXNAL, 66, 472 (1944). (2) Parts of this paper are taken from a thesis submitted by Leonard B. Chandler to the Graduate School of the Massachusetts Institute of Technology in partial fulfillment of the requirements for the degree of Doctor of Philosophy in October, 1939. A similar paper was read before the Division of Organic Chemistry at the Cincinnati meeting of the American Chemical Soicety in April, 1940. (3) Present address: Nylon Division, E. I. du Pout de Nemours and Company, Wilmington, Delaware. (4) Wohl, B n . , 16, 730 (1893): cj. Wohl and List, ibid., SO, 3101 (1897); Wohl, ibid.. SI, 3668 (1899). (5) Jacobi. ibid., IC, 696 (1891); rf. Rischbieth, ibid., IO, 2673 (1887). (6) Imine and Gilmour, J . Chem. SOC.. DS, 1429 (1908); 6f. (IS,622 (1931). Wolfrom and Thompson, Tms JOURNAL, (7) Brigl, Mtihlschlegel and Schinle, Be?., 84, 2921 (1931). (8) Hockett. TEXSJOURNAL., 67, 2265 (1935); Hockett, Deulofeu, Sedoff and Mendive, ibid., SO, 278 (1938). (9) Deulofeu, J . Chum. SOC.,2973 (1932). (10) Hockett, Dienes, Fletcher and Ramsden, TEISJOUXNAL. 6S, 167 (1944).

It was shown by Maquenne“ that silver ions are not necessary for the removal of the cyanide group and that aqua ammonia alone will bring about the reaction in one instance. In this Laboratory Maquenne’s observation has been confirmed and the effectiveness of 29% ammonia alone has been shown in a number of other cases.8 Interesting questions arise concerning the mechanisms of the several reactions which convert pentaacetyl-D-gluconic nitrile into “D-arabinose diacetamide.” It is noteworthy that no tendency of free sugars to condense directly with acetamide has been observed either in this Laboratory or elsewhere’ in spite of a number of efforts to effect such a combination under various conditions. The implication is that some form of sugar other than the pyranose is present a t the instant of combination with acetamide. Such a form of D-arabinose could conceivably be the tetraacetyl aldehydo-D-arabinose which might be present in a transitory manner, ij the hydrolysis of nitrile group proceeds a t a greater speed than ammonolysis of the acetate groups. The latter, so long as present, will prevent cyclization to furanose or pyranose. C=N

I I

/OH HCOAc

HCOAc

I

AcOCH

1

NH4OH ACOCH

-

HCOAc HLOAc

I

HICOAC

1

-

HC=O -HOAc

1

I

AcOCH

HCOAc

HCOAc

HCOAc

H OAc

HiCOAc NH4CN

HlCOAc

I

I

+

c:I

There are two objections to this theory. First, we have been unable to combine pentaacetyl aldehydo-D-glucose with acetamide in this Laboratory under any conditions found so far. Second, in the Wohl reaction no acetamide is present until it has been formed by the action of ammonia upon acetate groups in the molecule. We must therefore assume that two acetyl groups are converted to acetamide prior to condensation. If acetyl groups must also serve to prevent cyclization until after condensation has occurred, we must also assume that the groups at the two and three positions undergo ammonolysis more rapidly, than those in the four and five positions. Thus 4,5-diacetyl aldehyde-D-arabinose may be the true intermediate (11) Maquenne, Compr. rrrtd., 1S0, 1402 (1900)

ROBERTc. HOCKETT AND LEONARD B. CHANDLER

958 HC=O

HC-(NHCOCH,)z

I

HOCH

I

HCOH +ZCH&ONH, HCOH I + I H~OAC - H ~ O A ~

I

2”s

- I

I

HICOAc

ported formulation as an Nacetyl-D-glucosylamine. Such a compound had previously been reported by Brigl and Keppler HCOH and by Niemann and Hays.I5 HCOH Their product, prepared by the HpCOH I action of ketene upon D-glucopyranosylamine, was shown by oxidation with periodic acid and hypobromite to a dibasic acid with one less carbon atom. to contain a pyranose ring. COMPARISON O F N-ACETYL-D-GLUCOSYLAMINES HC(”COCHs)2

HOAH

I

Vol. 66

HlCOAc

I HOCH I

The non-combination of the fully acetylated aldehydo sugars with acetamide may indicate either (1) that removal of acetyls from positions two and three activates the carbonyl in a necessary manner or, (2) that an alkaline medium is Melting point, necessary for the condensation. If this is true, ‘ C . (uncor ) Specific rotation the “sugar diacetamide” is obviously formed as a Hockett and Chandler 192-194 +86.9 (C, 1.584; result of a delicate balance among the rates of Hz0; 25”) several concurrent reactions. Kiemann and Hays -22.4 (C, 1.24; In either case, fully acetylated aldelzydo-sugars (Brigl and Keppler) 255 HzO) ought to be converted by aqua ammonia into the COMPARISON OF PENTAACETYL-D-GLUCOSYLAMINES open-chain diacetamide condensation products. Deulofeug obtained erythrose diacetamide” Hockett and Chandler 82 5-84 5 $32.7 (C, 1.062; CHCls; 25 ”) by the action of ammonia in ethanol upon tri$17.7 (C, 0.76; acetyl-aldehydo-L-erythrose and Brigl, Miihl- Niemann and Hays (Brig1 and Keppler) 160-161 CHCla; 22”) schlegel and Schinle7 obtained an analogous I‘D-glucose dibenzamide” by the action of To determine the nature of the ring in our isoammonia in methanol upon pentabenzoyl-alde- mer, we applied the lead tetraacetate oxidation hyh-D-glucose. An attempt in this Laboratory, method.16 Formaldehyde was produced and to prepare “D-xylose diacetamide” has resulted an oxidation curve of type IV was obtained in failure. The action of aqua ammonia upon (Fig. 1). The substance is therefore considered crystalline tetraacetyl-aldehydo-D-xylose,lzhas to be an N-acetyl-D-glucofuranosylamine. It is yielded no crystalline products. Neither have impossible to judge at present whether this S~ or penta- product is formed by loss of acetamide from a we been able to obtain D - ~ U ~ Ooxime acetyl-D-gulonic nitrile in a crystalline state. typical “hexose diacetamide”” or by direct conSince “0-glucose diacetamide” had apparently densation of a glycofuranose with acetamide. not been described, we undertook to prepare this HCOH HC-NHCOCH, substance by two methods: the action of 29% I\ aqua ammonia upon pentaacetyl-aldehydo-D-gluHA& HCOH I cose13and by the action of the same reagent upon hexaacetyl-D-a-glucoheptonic nitrile14 prepared from D-a-ghcoheptose oxime. In both cases HC--J’ , HC-J the same crystalline product, melting 192-194’, I was obtained in yields of 56 and 26% of the HC0.4c HCOH theoretical, repectively. Nitrogen analysis I HZCOAc HKOH showed a t once that the substance was not a typical “aldose diacetamide.” The analysis supWe describe D-a-glucoheptose oxime apparently * ” for the first time. It was previously prepared in this Laboratory by Mr. AndrC Laus (unpublished research) in 1036. We wish to thank Professor M. L. Wolfrom of The Ohio State University for seed crystals of three compounds used in this study and Dr. Morris Zief of this Laboratory for optical measurements on glucoheptose oxime. Experimental 0

20

.______

40

60

80

100

120

140

Time in hours. Fig. 1.

(12) Wolfrom, Newlin and Stahly, THISJOURNAL, SS, 4379 (1931) (13) Wolfrom, rbrd., 61, 2188 (1929). (14) Cf.Zempl6n and Kiss, Ber , 60, 165 (1927)

N-Acetyl-D-glucofuraosylamine from Pentaacetyl-aldehydo-D-glucose.-i sample of 6 9 g. 10.0177 mole) of (15) Brigl and Keppler, Z . physiol Chcm., 180, 38 (1929); Niemann and Hays, THISJOURNAL, 63, 2060 (1940). (16) Hockett, Reeder and Nickerson, ibid., 66, 472 (1944). ( 1 7 ) Cf. Schneider, Sepp and Stiehlcr, Bcr., Sl, 220 (1918); Pacsu, ibid.. 68, 509 (1925): Montgomery, Hann and Hudson, THIS J O U R - A L , 89, 1121 (1937)

June, 1944

STRUCTURE OF N-ACETYL-D-GLUCOFURANOSYLAMINE

pentaacetyl-aldehydo-D-glucose(m. p. 116-118'), prepared by the method of Wolfromla was treated with 110 cc. of 29% aqua ammonia at 60-60" for three hours. The solvents were then removed under reduced pressure and the remaining sirup was treated again with 110 cc. of ammonia solution. After concentration to a sirup, the latter was treated with 50 cc. of absolute ethanol which caused separation of fine crystals weighing 2.2 g.; yield, 56%. Recrystallized from about five parts of water with addition of eighteen p ts of absolute alcohol, the crystals reached a maximu@%melting point of 189-191 a (uncor.) (with decomposition) when heated a t the rate of two degrees a minute. They rotated's $86.7' (C, 3.937; Hz0; 22.4'). A slow mutarotationl6 to 4-85.8" occurred in ten days. The compound is moderately soluble in hot water, slightly soluble in ice-cold water and in hot dry methanol, and insoluble in absolute ethanol, ether, pyridine, acetone and chloroform. It reduced Fehling solution only slowly on continued boiling. Anal. Calcd. for CaHlbOsN: C, 43.3; H, 6.84; N, 6.34. Found: C, 42.4,42.2, 43.2; H, 6.39, 6.73, 6.26; N, 6.34, 6.49. Pentaacetyl-D-glucofuranosylamine.-A sample of 1.56 g. (0.00707 mole) of N-acetyl-D-glucofuranosylamine was treated with a mixture of 7.6 cc. (0.0789 mole) of aceti: anhydride and 6 cc. (0.0740 mole) of pyridine a t 75-80 for three hours. Solution was complete in thirty minutes and then slight discoloration occurred. After three hours, 5 cc. of pyridine was added and the solution was concentrated under diminished pressure (bath temp., 55') to a sirup. This sirup was several times dissolved in sbsolute alcohol and reconcentrated to get rid of pyridine. The resulting yellow sirup was dissolved in 10 cc. of acetone and filtered through carbon to decolorize. The filtrate was treated with 10 cc. of dry ether and then with petroleum ether to turbidity. Crystals weighing 1.46 g. (53% of the theoretical) separated in a refrigerator. These pnsms, once recrystallized from the same solvents, melted 82.5-84.5" (uncor.) and rotsted'a +32.7' (C, 1.620;CHCls; 25'). They are readily soluble in acetone, chloroform, methanol and ethanol, somewhat soluble in warm water, very slightly soluble in boiling ether or cold water, and insoluble in petroleum ether. Anal. Calcd. for CMHSOI~N: N, 3.60; acetyl, 44.3. Found: N, 3.76, 3.77; acetyl, 44.2. Pentaacetyl-D-glucofuranosylamine was also made by heating at 90' for ninety minutes with four parts of acetic anhydride and one-quarter part of anhydrous sodium acetate. The mixture was poured on ice, and the product was dissolved in chloroform, which was shaken repeatedly with sodium bicarbonate solution, was washed with water, dried over sodium sulfate and evaporated. The sirup remaining was crystallized as described above. Attempted Condensation of Pentaacetyl-abhhydo-Dglucose with Acetamide.-After ten hours of boiling under reflux, a solution of the glucose derivative and of acetamide in chloroform was found to contain the starting materials unchanged. D-WGlucohaptose Oxime.-Nineteen grams (0.0906 mole) of D-a-glucoheptose (m. p. 188-190°) made as described by Haworth, Hirst and Stacey,lB was treated with a methanol solution of 0.129 mole of hydroxylamine overnight a t room temperature. A little unreacted sugar remaining the next morning was dissolved by warming a short time on the steam-bath. The solution was then concentrated to a sirup (diminished pressure) and was dissolved in a minimum volume of dry methanol. A crystalline product weighing 17 g. (85% ff the theoretical) soon separated which melted a t 89-91.5 . Recrystallized from water, it melted at 1oO-10lo(uncor.). (18) All rotations in this paper refer to specific rotations of the D line of sodium at 20' unless otherwise stated. Concentrations are stated in grams of solute in 100 cc. of solution. (19) Haworth, Hint and Stacey, J . Chum. Soc., 2864 (1931); cf. Fischer, Ann., M O , 64 (1892), and Philippe. Ann. ckim. phyr.. I6,289 (1912).

859

A d . Calcd. for C7HI6O7N: N, 6.22. Found: N, 6.33, 6.30. The substance shows a slight mutarotation in water. The initial rotation18 of -6.3" (C, 1.589; H20; 22") reaches 0.9 ' in seventy hours. Because of the low magnitude of these rotations, no accurate rate constant has been obtained. Hexaacetyl-D-a-glucoheptonic Nitrile.-Thirteen grams (0.0578 mole) of D-a-glucoheptose oxime (m. p. 89-91') was treated with 69 cc. of acetic anhydride and 5 g. of anhydrous sodium acetate. Heat was evolved as the oxime dissolved. Finally the mixture was digested for an hour a t 100" and then poured on ice. The precipitated gum soon solidified after excess acetic acid had been neutralized with sodium bicarbonate. Filtered, decolorized, and recrystallized from hot ethanol, 13.5 g. of product (5lC&of the theoretical) was obtained. After recrystallization to constant properties, the elongated prisms melted 85.587.5' (uncor.) and rotatedl8 $24.3' (C, 3.893; CHCl,; 24.8"). ZemplCn and Kiss, who prepared the compound by another method,'& reported a melting point of 1 1 2 . t 113.5' but a closely agreeing rotation +24.6" at 21 . Possibly two crystal modifications exist. Anal. Calcd. Cl~H46012N:C, 49.5; H, 5.44; N, 3.05. Found: C, 49.8, 49.8; H, 5.47, 5.47; N, 3.26. Acetyl determination by the method of Kunz and Hudsonz0resulted in hydrolysis of the cyanide group as well as the acetyls: calcd. for seven CHsCO, 65.6; found, 63.6 (60 min.), 64.7 (75 min.) and 65.6 (120 min.). N-Acetyl-D-%ucofuanosylamine from Hexaacetyh-aglucoheptonic Nitrile.-A sample of 10.5 g. (0.0229 mole) of the nitrile was heated for three hours a t 50' with 125 cc. of 29% aqua ammonia. The solution was then concentrated to a sirup (diminished pressure) which was given a second treatment with ammonia. The sirup obtained by a second concentration was dried by adding absolute alcohol and redistilling solvent-free. The dried sirup was dissolved in a little absolute ethanol and ether was added to turbidity. Crystals separated in a yield of 1.3 g. or 26%. Recrystallized as described above, the small prisms attained a melting point of 192-194" (uncor.) and rotatedI8 +%.go (C, 1.584; H20; 25.1'). The two samples when mixed showed no depression of melting point. AnaZ. Calcd. for CaHlbOeN: C, 43.4; H, 6.84; N, 6.34. Found: C, 43.4, 43.8; H, 6.94, 7.12; N, 6.25. Rate of Oxidation of N-Acetyl-D-glucofuosylamine by Lead Tetraacetate.-The measurement was camed out under the standard conditions described by Hockett, Dienes and Ramsden." Detection of Formaldehyde on Oxidation of N-Acetyl-Dglucofuranosylamine by Lead Tetraacetate.-A sample of 0.10 g. (0.0005 mole) of the glucose derivative was dissolved in 10 cc. of water in a small distilling flask. A volume of 16 cc. of approximately 0.05 molar lead tetraacetate in glacial acetic acid was added from a buret. After ten minutes, a receiver was attached and packed in dry-ice. The liquids were distilled under reduced pressure into this receiver. To the distillate was added 30 cc. of a saturated solution of 2,4-dinitrophenylhydrazine in 2 N hydrochloric acid (about 0.2 g. of the hydrazine). The precipitate of formaldehyde 2,4-dinitrophenylhydrazone melted a t 166' (cor.). summary

An N-acetyl-D-glucofuranosylaminehas been prepared in two ways: by the action of aqua ammonia upon pentaacetyl-aldehydo-D-glucose and upon hexaacetyl-D-a-glucoheptonicnitrile. The physical properties of the new compound are recorded along with those of the corresponding pentaacetyl-D-glucofuranosylamine and of ~ - a glucoheptose oxime. (20) Kunz and Hudson, THISJOURNAL, 'M, 1932 (1928). (21) Hockett, Dienes and Ramden, ibid., M, 1474 (1948).

RICHARDT.A R N O L D

'360

AND

The ring structure of N-acetyl-D-glucofuranosylamine has been determined by its behavior when oxidized by lead tetxaacetate. The new

[CONTRIBUTION FROM THE

RODERICK A. BARNES

Vol. (iG

reactions give further insight into the mechanism of the Wohl degradation of sugars. CAMBRIDGE,

MASS.

RECEIVED FEBRUARY 23, 1944

SCHOOL O F CHEMISTRY OF TEE UNIVERSITY OF M I m S O T A ]

The Jacobsen Rearrangement. VIII.' Cyclic Systems; Mechanism BY RICHARD T. ARNOLDA N D RODERICK A. BARN&

For many purposes it is convenient to regard hydrindene and tetralin as o-dialkylbenzenes. If this view is accepted, then one might suppose that the tri- and tetramethylene rings in these substances would undergo rearrangement when the molecule is subjected to conditions which bring about a migration of simple alkyl groups. Therefore it was of interest to study the Jacobsen rearrangement of hydrocarbons derived from tetralin and hydrindene. During the course of this investigation octahydroanthracene (I), 5,6,7,8-tetrahydrobenz[f]indan (III),s 6,7-diethyl-

R = oxidizable side chain of unknowii structure

tar

+ starting materiel

X ~ 1 ~ 1 ~ 4

R

R

I R

I

COOH

I

XI

I11

IV

COOH _c HOOC@COOH COOH

RI

CJ%

HOOC/

XIS

VI1

VI

'

HnSOdt

C O :XIIl : - C H t AlCIzi

V

R AIClsi

CzHs

GH5

cb\c2Hk VI11

I

I

Pd-C_

0 3 \ C z H s IX

(1) VII, THISJOURNAL, 69, 2631 (1940). The Authors desire to thank Dr. Lee Irvin Smith for his interest in this work and for the authentic samples of benzene tetracarhoxylic acids used as reference compounds. (2) Abstracted from athesis by Roderick A. Barnes, presented to the Graduate Faculty of the University of Minnesota in partial fulfillment of the requirements for the Ph.D. degree, Novemher, 19-13 (3) Sen-Gupta. 1.I n d . Chcm. Soc.. 16, 89 (1939).

+ XI11

(R =

CHs

or C2Hb)

tetralin (V) , 5-ethyl-6-methylhydrindene(XI), and S-hYdrindacene (x)(Ring Index 1459) have been prepared and their reactions with sulfuric acid and aluminum chloride (in three cases) investigated, The only example of a Jacobsen rearrangement involving a cyclic system which so far has been